arXiv 2602.16710
EgoScale: Scaling Dexterous Manipulation with Diverse Egocentric Human Data
By Ruijie Zheng, Dantong Niu, et al.
Published 2026-02-18
Mindmap
Browse the paper's core ideas, clusters, and relationships in a structured outline.
Human behavior is among the most scalable sources of data for learning physical intelligence, yet how to effectively leverage it for dexterous manipulation remains unclear. While prior work demonstrates human to robot transfer in constrained settings, it is unclear whether large scale human data can support fine grained, high degree of freedom dexterous manipulation. We present EgoScale, a human to dexterous manipul…